Has anyone ever compiled stats on Board usage, with a breakdown of the individual forums by days of the week? I think it’d be interesting to see, for example, what day of the week the most Pit threads are started; when folks get into Great Debates most; etc, etc.
(I’m not talking about time of day, since we have Dopers worldwide, some people only access from work, and so on).
Just my dos centavos…
Do a serach for poster kabbes in this forum - I know he did a lot of work giving us just those stats - all without putting any strain on the server. I think this was before the Great Crash, but some of it may have been after, so it might be lost. Search isn’t going too well at the moment, but I did manage to get this one:
on Most-Active posters (from 2000) http://boards.straightdope.com/sdmb/showthread.php?threadid=29678
peepthis - the Straight Dope doesn’t publish statistics of that kind. Some posters have gathered a few numbers here and there but there is no organized effort. And we prefer that posters not do this on their own due to our current performance problems.
Of course, if you wanted to, for example, write down the number of threads per forum on the home page once a day or once a week, that would probably be pretty harmless. But we try to discourage any application doing automated calls to the SDMB webserver.
You might like to click the link in my sig.
pan
Y’know, kabbes, this is actually the first time that I’ve taken a close look at your formula, but is that dot between the t and the 2.3 a multiplication sign? Because if so, then the limit of your formula as t goes to infinity is 1, suggesting that everyone here sticks around forever.
[semi-hijack, part the first]
Are we back at the same traffic level we had before the crash?
[/semi-hijack, part the first]
[semi-hijack, part the second]
I sort of miss the line on the old home page that said “There are currently XXX registered members logged in and XX guests.” It was nice to see how many people were on the board at any given time.
[/semi-hijack, part the second]
Yeah Chronos - if you look around, you’ll actually find unpublicised threads in which I admit that my formula has a usable range of about 15 months. That’s all the data I had - anything beyond this date would have lacked too much credibility.
It actually works surprisingly well for the first 15 months or so though.
One day (a man can dream, can’t he?) I’ll have the data to construct the Grand Unified Theory of Dope. Sadly this is not that day.
pan
kabbes-
How does that formula work for the first month? Because I would think that there’d be a sharp drop within the first week. Or maybe you don’t have any data from then, either. Couldn’t you get a simpler formula, maybe just an exponential function, to approximate kabbes’ hypothesis on the 2-15 month range? Better make that 2-14 months, because it looks like your formula predicts the number of active dopers goes up in the 15th month.
Yeah, I should try to simplify the formula really. I just picked up the result from the model and dumped it into a post - something that I wouldn’t do IRL (but I’d say you got your money’s worth ;))
And yes, I have to admit that sub one month it isn’t going to be reliable either. Gee, way to destroy the analysis guys 
Now let us hold hands and pray to the gods of data (Paradox, Access and SQL) for another bountious gift. They may someday favour us - you never know.
pan
Of course, if you’re going to try to extend kabbes’ Formula to the first month, you’re going to have to ask whether it refers to “time since registration” or “time since first post”. Remember, over a third of our registered members have never posted, so it’s a worthwhile distinction.
What percent of thread hits are non members?
Are we singing to the choir?
Tris
FWIW, I did my analysis based on time since registration.